Resumen

GIRON-RIOS, Yolanda et al. Reflectance fractal analysis of saline soils. Agrociencia [online]. 2009, vol.43, n.4, pp.403-416. ISSN 2521-9766.

Variability in the soil salinity space has been documented in various studies. Reflectance is an essential variable addressed in them. In this study an alternative method is proposed for the analysis of salinity through the multiscalar data of reflectance. The suggested method extracts information to describe the soil structure as of the analysis of the soil spectral signatures using the fractal dimension (D). The structure of salt crusts was characterized as of the analysis of fractal signatures (FER) of reflectance spectra by using the techniques of rescaling (D_R/S) and wavelets (D_W) ranges. For the statistical analysis of FER the multivaried analysis was applied and the most significant relationships were identified between FER and the soil physical and chemical properties. Five hierarchical groups were obtained with the following order: g1< g2< g3< g4< g5, on the basis of their spectral responses evaluated on the wavelength of visible light (VIS) and near infrared (NIR) of the electromagnetic spectrum (EM). A comparative test was conducted taking into account the dimensions D_R/Sand D_W, which showed that the D_{R/ S} dimension for FER is a better indicator of the salt crust structure than D_W. The values of the D_R/Sdimension for each group exhibit the order 1.73> 1.70> 1.67> 1.66 for the groups g1, g2, g3, g4; and 1.67 for g5. A significantly high relationship was detected among the D_R/S dimensions when comparing the five groups. The D_W dimension showed no significant relationship with the morphology of crusts. Therefore, D is useful to classify the structure of salt crusts as of the soil fractal signatures (FER) extracted from the reflectance spectra.

Palabras llave : Fractal dimension; invariance to scaling; reflectance; salinity.

        · resumen en Español     · texto en Español     · Español ( pdf )